Nano-scale resolution X-ray computed tomography 1 Shawn Litster, Ph.D. Associate Professor, Department of Mechanical Engineering Carnegie Mellon University [email protected], 412 268 3050 Students: Siddharth Komini Babu, Sarah Frisco, Pratiti Mandal, William Epting, Arjun Kumar, Tim Hsu, Alex Mohammed Collaborators: Paul Salvador, Jay Whitacre, Ryan Sullivan, and Jessica Zhang (CMU); Hoon Chung and Piotr Zelenay (LANL) Workshop on 3D Microstructural Studies Carnegie Mellon University, July 8-10, 2015
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Nano-scale resolution X-ray computed tomography
1
Shawn Litster, Ph.D.Associate Professor, Department of Mechanical Engineering
Students: Siddharth Komini Babu, Sarah Frisco, Pratiti Mandal, William Epting, Arjun Kumar, Tim Hsu, Alex Mohammed
Collaborators: Paul Salvador, Jay Whitacre, Ryan Sullivan, and Jessica Zhang (CMU); Hoon Chung and Piotr Zelenay (LANL)
Workshop on 3D Microstructural StudiesCarnegie Mellon University, July 8-10, 2015
Outline
1. X-ray CT and nano-scale resolution X-ray CT (nano-CT)
2. Application of nano-CT to low temperature fuel cells for vehicles
3. Application of nano-CT to Li-ion batteries
2
Micro/Nano X-ray Computed Tomography
Slide 3
X-rayoptics
Micro X-ray CTConical beam focus~1 µm resolution
Nano X-ray CTX-ray optics50 nm resolution
Synchrotron light souce X-ray CTParallel, high intensity beamResolution of 15 nm with optics, limited depth of focus
Bone Photonic crystal Scaffold Shale RockIntegrated circuits
Images: Xradia.com
Why X-ray CT?• Three-dimensional: Important for anisotropic
materials (Compare to SEM or TEM)
• Internal imaging: Reveal inner structure within otherwise opaque materials (Compare to SEM and AFM)
• Non-destructive: Same sample can be imaged multiple times in different imaging modes and following various physicochemical events (Compare to destructive FIB-SEM).
• Ambient and controlled environments: No vacuum required, enabling in-situand in-operando experiments (Compare to SEM and TEM)
• Large field of view CT: Larger samples and more representative statistics (Compare to TEM-CT)
Slide 4
X-ray Attenuation and Radiographs
5
• Radiographs created by sample attenuating the X-ray
• Two main mechanisms for X-ray attenuation– Photoelectric absorption (dominant for 8 keV nano-CT)
– Compton scattering (hard X-rays)
• Photoelectric absorption scales as ~Z4/E3
• Beer’s Law
expo i i
i
I I x
Attenuated intensity
Initial intensity
Linear attenuation coefficientf(Z, density)
Thickness
Summation for material in series
http://www.nist.gov/pml/data/xraycoef/
Pt (Z = 78)
C (Z =6)
/r
[cm
2/g
]
Density normalized /r
Hand of W.C. Rӧntgen’s wife
Photon energy [MeV]
ThickerDenserHigher Z
More attenuation
X-ray Computed Tomography (CT)
6
• Transmission images collected over 180o
rotation with collimated beam
• Sample rotates in nano-CT
• 2D example for 4 radiographs
• Detector image intensity a function of sample density, atomic number Z, and thickness
Detector
Source
Filtered Back Projection Reconstruction
7
• Radiographs are computationally back projected into sample space
• High pass filtered to remove low frequency blurring
• Typically 400-900 projections for high resolution
Slice
Stacked slices 3D Volume
Nano-CT at Carnegie Mellon
Slide 8
• Xradia, Inc.’s UltraXRM-L200 Nano-CT
• Highest level resolution outside of a synchrotron using proprietary optics
• Laboratory 8 keV Cu rotating anode X-ray source
• Non-destructive imaging in ambient and controlled environments
• 4D imaging (space and time) for material evolution studies
• Fluid phase distributions
• 16 nm voxels, 50 nm resolution
NSF MRI award 1229090PI: Litster; Co-PIs: De Graef, Fedder, Feinberg, Sullivan
Nano-CT with X-ray Optics
Slide 9
Micro X-ray CT
Optics needed for resolutions better than 500 nm in X-ray CT
X-ray Source: Rotating Cu Anode• X-ray energy depends on target material’s emission
lines
• Trade-offs in photon energy:• Too low, no transmission
• Too high, low absorption contrast
• 8 keV of Cu anode balances absorption contrast and X-ray penetration depth for high resolution
• In between soft and hard X-rays
• Rotating anode for higher power
Slide 10
Emission lines of lab Cu X-ray source
X-rays
Monocapillary Condenser• Elliptical capillary focuses X-rays on sample
• High efficiency mono-capillary condenser instead of Fresnel zone plate condenser
• ~90% efficiency to enable high resolution with low intensity lab sources versus synchrotron beamlines
• Some inherent filtering above 10 keV
Slide 11
Fresnel Zone Plate Objective
Slide 12
• Diffractive X-ray lenses
• Circular grating with varying pitch focuses X-rays by constructive interference
• Rayleigh resolution depends on width of outermost zone
• For DRn = 35 nm, d = 43 nm
• Depth of focus (DOF) of ±16 µm for 8 keVphotons (λ = 0.15 nm)
• Optimized for 8 keV
• Long working distance (f) of ~2 cm
35 nmnRD
Gold rings
Inside the Nano-CT
Slide 13
The Optics – Two Modes
Slide 14
• M Feser, et al. Meas. Sci. Technol. 2008, 19, 094001
Pin hole Phase ring
Fresnel zone plate
Visible light microscope High Resolution (HRES):
Field of view: 16.3 µmResolution: 50 nm
Pixel resolution: 16 nm
Large Field of View (LFOV): Field of view: 65.6 µm
Resolution: 100-150 nmPixel resolution: 64 nm
• Each resolution mode has own set of optics (2 phase rings, 2 FZP, etc.)• Automated switching between LFOV/HRES and absorption/phase contrast
Resolution Test Pattern Images
Slide 15
Large Field of View – Phase Contrast High Resolution – Absorption Contrast
• Higher signal to noise ratio for pore/solid interface• Better resolved fine features of layered graphite• Little contrast between bulk solid and bulk air
Physics-Based Phase Contrast Correction Algorithm• Developed new physic based correction for Zernike phase
contrast to remove artifacts by modeling diffraction optics of the system
• Transforms phase contrast image to equivalent absorption image
18
Uncorrected Corrected
Correction applied to PTFE coated carbon fiber for PEFC gas diffusion layer
Kumar, Mandal, Zhang, and Litster, J. Applied Physics, in press, 2015
Radiographs and Reconstruction• Combustion particle analysis by Prof. Ryan Sullivan & Kyle
Gorkowski, CMU
• Particles captured on Kapton mesh sample mount inside inertial sampler
• Mesh directly transferred into nano-CT without further sample preparation
Slide 19
Kapton mesh imaged by nano-CT’s visual light microscope
Transmission Radiographs 3D Reconstruction of Particles in Field of View
Start
180o rotation of sample
Finish
Sample requirements• Sample must be small to fit within the field of
view during entire rotation (Not strict, but recommended for quality)• Max. dia. of 65 µm for LFOV
• Max. dia. of 16 µm for HRES
• Not an issue for thin-film (<5 µm)
– e.g., porous polycarbonate
• Ideal sample is a cylinder with diameter of the FOV width
• Sample must be stable• Hygroscopic materials can be an issue due to swelling
• Software can correct for some rigid body motion with gold fiducial particle.
Slide 20
Shearing et al., J.Euro.Ceram. Soc., 2010
Sample mounting
Slide 21
Clip type sample holder
Flat back sample holder Pin vise sample holder
1
32
Sample Preparation• Sample mounting ranges from very simple (film clip )
to challenging (FIB)
• Majority of samples are hand prepared at CMU and at Xradia’s applications lab
• Some options:• Microtome to thin polymer samples
• Hand cut sample to triangle so tip fits in FOV
• Laser milling
• Epoxy small and flimsy samples to pin
• Focused ion beam lift out
• Gold fiducial particles for HRES reconstruction
Slide 22
Catalyst particles on Kapton cut to fine tip
FIB lift out of SOFC electrode
Laser milled cylinder of shale rock
Setup for mounting gold fiducial particles
Sample Preparation – Miro Laser Milling
23
• Laser micromill with 1 µm accurate control and beam width to prepare sample columns from MEAs
• Prepare through-thickness samples with maximum sample size in 3D image
Hyundai ix Tucson FCEV• 100 kW PEFC• Li-ion battery (21 kW)• Range: 650 km at 70 MPa• Mileage: 72 mpgge combined
• High efficiency, long range, low emission fuel cell vehicles• Low temperature, high power density fuel cell• Acidic polymer electrolyte membrane• Pt and Pt-alloy catalyst
Toyota Mirai FCEV• 114 kW PEFC• Range: 650 km at 70 Mpa• 3 min refueling time• Mileage: 72 mpgge combined
Gas Diffusion Layer(GDL)
Pt catalyst
Carbon support
3-5 nm
40 nm H2OH+
O2
e-Agglomerate
Nafionbinder
Pores• Litster and McLean, J. Power Sources, 130, 61-76, 2004.• Uchida et al., J. Electrochem. Soc, 142, 4143-4149, 1995.
• Orange volume rendering shows high Z Fe from absorption contrast scan.
Effect of Nafion Loading on 1 atm Air Performance
30
• Low Nafion loading reduces activity Lower voltage at low current
• High Nafion loading increase transport loss Large voltage drop at high current
0 1 2 3 4 5 60
5
10
15
20
Diameter [(m)]
Vo
lum
e P
erc
en
tag
e [%
]
CM-PANI-Fe 35wt% Nafion
CM-PANI-Fe 50wt% Nafion
CM-PANI-Fe 60wt% Nafion
0.05 0.1 0.15 0.2 0.25 0.3 0.350
2
4
6
8
10
12
14
Diameter [(m)]
Vo
lum
e P
erc
en
tag
e [%
]
CM-PANI-Fe 35wt% Nafion
CM-PANI-Fe 50wt% Nafion
CM-PANI-Fe 60wt% Nafion
Hierarchical Morphology Properties
31
• Wide variety of morphological properties characterized, including effective length scales• Characterization effective diameter of various electrodes for pore and solid domains